It is common in inventory theory to consider policies that minimize the expected cost of ordering and holding goods or materials. Nevertheless, the realized cost is a random variable, and, as the Saint Petersburg Paradox reminds us, the expected value does not always capture the full economic reality of a decision problem. Here we take the classic inventory model of Bulinskaya (Theory of Probability & Its Applications, 9, 3, 389–403, 1964), and, by proving an appropriate central limit theorem, we show in a reasonably rich (and practical) sense that the mean-optimal policies are economically appropriate. The motivation and the tools are applicable to a large class of Markov decision problems.
Inventory management Markov decision problems Central limit theorem Non-homogeneous markov chain Dobrushin coefficient Stochastic order Discrete-time martingale
Arlotto A, Steele JM (2016) A central limit theorem for temporally non-homogenous markov chains with applications to dynamic programming. Math Oper Res 41(4):1448–1468MathSciNetCrossRefMATHGoogle Scholar
Bulinskaya EV (1964) Some results concerning optimum inventory policies. Theory of Probability & Its Applications 9(3):389–403CrossRefMATHGoogle Scholar
Li X, Shaked M (2007) Stochastic orders based on total time on test transform. In: Advances in statistical modeling and inference, Vol. 3 of Ser. Biostat., World Sci. Publ., Hackensack, NJ, pp 131–143Google Scholar